All-zero block detection in VC-1

VC-1 is a video codec specification that has been standardized by the Society of Motion Picture and Television Engineers (SMPTE) and implemented by Microsoft as Windows Media Video (WMV) 9. In this paper, zero-block detection is used to reduce the computation in the VC-1 encoder for very low bit rate video, like videophone applications. The zero-block detection depends on comparing the sum of the absolute difference (SAD) of a residual block to a value related to the quantization parameter (QP) during the process of the motion search for the enter-frame encoding. We introduced two options. One is used to save additional time and the other one is used to preserve the quality. One of the VC-1 innovations is how it deals with an entire block of zero coefficients. The transform information is not sent, and the inverse transform is a zero block. Preserving the quality, results showed that an adaptive technique can reduce the encoding and decoding times, up to %3.5 in encoding and %10.5 in decoding times, and the code length, up to %8 reduction. This comes with negligible reduction in the quality represented by the peak signal to noise ratio (PSNR), less than 0.2 db. The measurements based on the SSIM, structural similarity shows improvement in the perceptual quality of the decoded sequences.